Minimum number of turns needed for an active pickup?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
What gauge of enamelled copper wire? What size pole? What type and size of magnet, neodymium or ferrite?

As for the specifics, I have no idea. The thin gauge wire on conventional pickups is extraordinarily fragile so I was hoping to use something thicker that of course would mean having to reduce the number of turns.

600 to 1800 sounds like a good range compared to the 7000 - 8000 used in a passive pickup.
 

PRR

Member
Joined 2003
Paid Member
If you use a high-OSI amp like the popular TL072 (or 12AX7), then standard turn-count is needed for generally satisfactory hiss performance. Putting the first stage in the guitar does not magically change things.

A medium-OSI amp like '5532 working with a little gain can probably do OK with a lower turns-count, in the area you are proposing.

Note though that a significant part of pickup "tone" is self-inductance resonating against cable capacitance, the top-ringing. Lower N is lower L and your cable is about zero, so it is "flat" where conventional pickups have a peak.
 
Disabled Account
Joined 2010
Assuming you reduce the number of turns to 1/10 of the usual, impedance would drop by a factor of 100, i.e. from 100~500k down to 1~5k. There are numerous OP-amps available that provide low noise at these impedance like NE5532, NJM4580,OPA1602, OPA1612.. just to mention a few. Even a discrete bipolar amp could do a nice job here.

The missing peak can be reproduced by parallel caps.
All in all - these high number of turns are a historical remnant from the tube era and other solutions are possible. (for people with an open mind:D)
 

PRR

Member
Joined 2003
Paid Member
> reduce the number of turns to 1/10 of the usual, impedance would drop by a factor of 100

Yes; but output level drops to 1/10.

There are not a lot of opamps with noise voltage 1/10th of a TL072 or 12AX7. Many of the ones with super low Ehiss have high Ihiss, high enough to dominate even at the much lower impedance level.
 
Assuming you reduce the number of turns to 1/10 of the usual, impedance would drop by a factor of 100, i.e. from 100~500k down to 1~5k. There are numerous OP-amps available that provide low noise at these impedance like NE5532, NJM4580,OPA1602, OPA1612.. just to mention a few. Even a discrete bipolar amp could do a nice job here.

I have some AD620 and INA122 instrumentation op amps left over from another project. Would these work well for this kind of application?

http://www.analog.com/media/en/technical-documentation/data-sheets/AD620.pdf
http://www.ti.com/lit/ds/symlink/ina122.pdf
 
Last edited:
Disabled Account
Joined 2010
I have some AD620 and INA122 instrumentation op amps left over from another project. Would these work well for this kind of application?

http://www.analog.com/media/en/technical-documentation/data-sheets/AD620.pdf
http://www.ti.com/lit/ds/symlink/ina122.pdf


For my knowledge instrumentation amps are more noisy than simple Op-amps.
It is true, you have to consider voltage and current noise of you input amp as well.

NE5534/5532 for instance will give you a noise figure close to thermal resistor noise at source impedance between 2k and 5k.
All in all there is no theoretical difference in noise figure comparing hi-level hi-impedance P.U. with its lo-level lo-impedance counterpart.

Anyway the physical limits (at room temperature) are reached and then there is no way for further improvements.
 
Using less turns means that if you can use a balanced preamp, the need for screened leads going to the pick up then becomes a twisted pair. There is little or no noise then to produce hiss or hum etc.
Most good manufacturers use the TL061 op amp with the input between pins 2 & 3, as it is quite happy on a 9volt supply; (+ - 4.5v) and it is a low noise device.
 

PRR

Member
Joined 2003
Paid Member
AD620 and INA122 instrumentation op amps ... TL061 ...

Yes, you can build a house with whatever wood you have in the yard or bring home.

None of these three would be a first-choice for "low hiss" in a medium-low impedance guitar pickup application.

The '062 is clearly inferior for hiss to stand-by '072 or historical reference 12AX7. It may be 'popular', but that may be price, battery consumption, and general don't-care attitude.

Classic 5000+turn pickups work well with input hiss levels like 2uV. If you go down to 1000 turns you have 1/5th the output level. To get back to normal level you need gain of 5 (easy). To not have "higher hiss" in this new design you must have 1/5th the amplifier hiss of the old standard amps. This means knowing how to navigate the specs. The obvious jellybean for ~~1/5th the hiss voltage of TL072 is the 5532. Better is possible, though not easy.

Balanced can reduce external crap. However so can shielding. There is a practical limit: the pickup can never be perfectly shielded. And it is difficult to wind a pickup perfectly balanced; the inside and outside of the winding are not equally exposed to external crap.

Balanced does NOT reduce the intrinsic self-hiss of ALL real sources. All resistors hiss. All sources (and amplifiers) have resistance.
 
Last edited:

PRR

Member
Joined 2003
Paid Member
> couple of dB quieter

Have you confirmed this?

Across the several things now sold as 553*?

My feeling has always been that the single and dual tend to the same hiss level. Putting a higher number on the marketing for the dual reduces rejects and justifies paying more/amp for the single.
 
I have some AD620 and INA122 instrumentation op amps left over from another project. Would these work well for this kind of application?
A lot depends on what sort of guitar sounds you're going for. Clean or slightly distorted, relatively low gain? You probably won't encounter hiss issues.

It's the metalheads with their uber-distortion, ultra high gain amps that struggle to manage hiss.

A few years ago, one of the well known guitar/pickup manufacturers was trying to sell active guitar pickups with printed coils (relatively few turns printed on PCB) rather than traditional windings. I remember the demos sounded too bright with too much HF response to sound "right".

-Gnobuddy
 
Administrator
Joined 2004
Paid Member
> couple of dB quieter

Have you confirmed this?

Across the several things now sold as 553*?

My feeling has always been that the single and dual tend to the same hiss level. Putting a higher number on the marketing for the dual reduces rejects and justifies paying more/amp for the single.

Last time I checked the 5534 was not exactly the single version of the 5532, but a slightly different chip. And yes I did measure them in a variety of different applications including phono pre-amplifiers and they generally proved to be several dB quieter. (I will admit this was two decades ago, but a variety of chip makers production were tested.)
 
A few years ago, one of the well known guitar/pickup manufacturers was trying to sell active guitar pickups with printed coils (relatively few turns printed on PCB) rather than traditional windings.
It was the Fishman Fluence pickups I'd been thinking of: Unwound: Fishman Rethinks the Electric Guitar Pickup | Premier Guitar

I don't know what type of devices were used to amplify the tiny voltage output from the printed coils up to traditional high-Z pickup levels, but I doubt it was anything more exotic than a couple of budget opamps.

-Gnobuddy
 
It was the Fishman Fluence pickups I'd been thinking of: Unwound: Fishman Rethinks the Electric Guitar Pickup | Premier Guitar

I don't know what type of devices were used to amplify the tiny voltage output from the printed coils up to traditional high-Z pickup levels, but I doubt it was anything more exotic than a couple of budget opamps.

-Gnobuddy

There's very little public technical information about these. I see figures of two coils (one dummy) each with 48 layers and an unknown number of tracks.
Battery life from a 9v cell is quoted at 200 hours which equates to a current draw of about 1.5mA which is much greater than the 80-220uA quoted for active EMGs which appear to use slightly underwound conventional coils.
 
Disabled Account
Joined 2010
to judge input voltage noise of opamp you must relate to actual source impedance.

A typical pickup with 5-10H inductance exhibits a peak impedance in the range of 500kOhm.
Noise density of a 500kOhm resistor is 80nV/sqrt(Hz).
TL061 with en=42nV/sqrt(Hz) is not a bad choice in that case.
Certainly this assumption holds only around the resonant peak, so some improvement with lo-noise FETs is possible.
Lowest noise voltage and current can be achieved using JFETs like 2SK170 - parts that are getting rare birds nowadays.

Some days ago I checked my guitar-pre-amp with a JFET-source follower input stage, followed by op-amps.
Due to finite drain resistance, the power supply rejection was insufficient in my application.
Remowing the source-follower and routing the input to the op-amp directly improved this significantly without degrading noise at all.
Op-amp was OPA1678 (CMOS input, 4,5nV/sqrt(Hz)).
 
Last edited:
coils...48 layers...an unknown number of tracks.
We can make an educated guess about the number of tracks using the minimum cheap, off-the-shelf spacing between PCB tracks that are available from budget-priced PCB manufacturers, combined with the available space on the PCB. Fluence pickups fit into the same footprint as stock Fender or Gibson pickups.

Without crunching any numbers, my first guess is that there are unlikely to be more than ten turns per PCB layer in the Fender-sized versions, so 48 layers will translate to roughly 500 turns. I will be surprised if this guess is not within a factor of two of the correct answer.

-Gnobuddy
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.